Very long inputs
Choose Grok 4.20 Non-Reasoning
Grok 4.20 Non-Reasoning reads up to 1M tokens per request, against 200k for Claude Haiku 3.5.
Claude Haiku 3.5 vs Grok 4.20 Non-Reasoning: price, context and capabilities
Choose Grok 4.20 Non-Reasoning for very long inputs and the latest release; on the other points the two are level or the data is missing.
In short
- Claude Haiku 3.5 has been retired; Grok 4.20 Non-Reasoning is the one you can still call.
- A price comparison is not possible: one of them has no comparable published price.
- Grok 4.20 Non-Reasoning reads more per request: 1M tokens against 200k.
- Grok 4.20 Non-Reasoning is the newer model (March 2026, against October 2024).
Which should you choose?
Each answer below comes straight from the published prices, specifications and independent benchmark results.
Latest release
Choose Grok 4.20 Non-Reasoning
Grok 4.20 Non-Reasoning is the newer release (March 2026, against October 2024 for Claude Haiku 3.5).
Specifications side by side
| Claude Haiku 3.5 | Grok 4.20 Non-Reasoning | |
|---|---|---|
| Provider | Anthropic | xAI |
| Category | Language models | Language models |
| Status | Retired | Generally available |
| Headline price | $1.60 / 1M tokens | $1.56 / 1M tokens |
| Input | $0.80 / 1M tokens | $1.25 / 1M tokens |
| Output | $4.00 / 1M tokens | $2.50 / 1M tokens |
| Cached input | $0.08 / 1M tokens | $0.20 / 1M tokens |
| Context window | 200k | 1M |
| Max output | No data | No data |
| Released | October 2024 | March 2026 |
| Knowledge cutoff | No data | No data |
| Licence | Proprietary | Proprietary |
| Capabilities | Vision, Tool calling, Structured output, Prompt caching, Batch API | Vision, Tool calling, Structured output, Prompt caching, Batch API |
| API model id | claude-3-5-haiku-20241022 | grok-4.20-0309-non-reasoning |
Benchmarks side by side
Leader in bold.
| Benchmark | Claude Haiku 3.5 | Grok 4.20 Non-Reasoning |
|---|---|---|
| Terminal-Bench Hard | No data | 16.7% |
| GPQA Diamond | No data | 77.6% |
| Humanity's Last Exam | No data | 27.9% |
| IFBench (instruction following) | No data | 49.3% |
| τ²-Bench (agentic tool use) | No data | 59.9% |
| MMMU-Pro (visual reasoning) | No data | 64.9% |
| Long-context reasoning (AA-LCR) | No data | 23.3% |
Independent evaluations by Artificial Analysis, run the same way for every model, at the reasoning setting shown. Percentages unless marked as an index.
Frequently asked questions
Should I choose Claude Haiku 3.5 or Grok 4.20 Non-Reasoning?
Choose Grok 4.20 Non-Reasoning for very long inputs and the latest release; on the other points the two are level or the data is missing.
Which is cheaper, Claude Haiku 3.5 or Grok 4.20 Non-Reasoning?
A price comparison is not possible: one of them has no comparable published price.
Which has the larger context window, Claude Haiku 3.5 or Grok 4.20 Non-Reasoning?
Grok 4.20 Non-Reasoning reads more per request: 1M tokens against 200k.
What is the difference between Claude Haiku 3.5 and Grok 4.20 Non-Reasoning?
Claude Haiku 3.5 is made by Anthropic and Grok 4.20 Non-Reasoning by xAI. Claude Haiku 3.5 has been retired; Grok 4.20 Non-Reasoning is the one you can still call. A price comparison is not possible: one of them has no comparable published price. Grok 4.20 Non-Reasoning reads more per request: 1M tokens against 200k. Grok 4.20 Non-Reasoning is the newer model (March 2026, against October 2024).
Which is newer, Claude Haiku 3.5 or Grok 4.20 Non-Reasoning?
Grok 4.20 Non-Reasoning was released on March 2026 and Claude Haiku 3.5 on October 2024.